Philippe Menasché is a cardiovascular surgeon at Hôpital Européen Georges Pompidou & University Paris V and Director, INSERM Research Unit 633, Cell Therapy in Cardiovascular Diseases, 20 rue Leblanc, Paris 75015, France. philippe.menasche@hop.egp.ap-hop-paris.fr
New evidence that human embryonic stem cells can rescue injured heart may bring these cells closer to clinical use.
Rapid advances in developmental biology over the last decade have continued to strengthen the rationale of regenerative medicine, which aims to repair damaged tissues with new cells mobilized from endogenous pools or supplied from exogenous sources. In the setting of cardiac diseases, this novel therapeutic approach has focused on heart failure because of its increasing prevalence, the limitations of contemporary treatments and the related financial bur-den. In this issue, Kehat et al.1 show that transplanted cardiomyocytes derived from human embryonic stem (ES) cells can substitute for pacemaker cells in a swine model of atrioventricular block and elicit an ectopic rhythm compatible with the animal's survival (Fig. 1). Their results provide compelling evidence that the graft integrated electromechanically within the recipient tissue.
Cardiomyocyte-like cells differentiated from human embryonic stem cells were transplanted into the left ventricular wall of pigs with complete heart block. The transplanted cells functioned to rescue cardiac rhythm.
Multiple cell types have been tested experimentally in animal models of myocardial infarction, with functional improvement as the primary end point. So far, safety and practicality considerations have dictated the use of autologous skeletal myoblasts2 and bone marrow−derived cells3,
4 as the first cell types to be tested in patients. Several trials have been completed or are underway. Their results, although often reported as positive, should be interpreted cautiously, as most of these trials have an usually small number of patients and lack randomization, double-blinding and placebo- controlled groups. Irrespective of the ultimate results, it now seems clear that skeletal myoblasts and bone marrow cells share a fundamental limitation: an inability to convert into true cardiomyocytes that could replace those irreversibly injured by heart attack(s).
This limitation has always been clear for myoblasts5, which, once engrafted, remain committed to their skeletal muscle phenotype. For bone marrow cells, the issue remains more controversial, although their alleged developmental plasticity6 has been seriously challenged by studies7,
8 that have used unambiguous genetic tracking methods to show that what was mistakenly interpreted as transdifferentiation may have corresponded to fusion events9 or to immunohistochemical artifacts. Even if skeletal myoblasts and bone marrow cells cannot transdifferentiate, they may still have functional benefits, mediated by a limitation of ventricular dilatation or paracrine induction of angiogenesis. Yet it is increasingly clear that these cells fail to satisfy the two major prerequisites for cardiac regeneration: an electrical coupling of the grafted cells with host cardiomyocytes and the subsequent generation of an active mechanical force.
The difficulty of inducing adult stem cells to cross their lineage boundaries gives new impetus to the intuitively appealing idea that the most appropriate cells for replacing dead cardiomyocytes might turn out to be cardiomyocytes. This view is supported by groundbreaking proof-of-principle experiments showing that transplanted fetal cardiac cells successfully engrafted into myocardial scars, connected with their host neighbors and improved function10. However, the ethical, availability, scalability and immunological issues associated with fetal material make it unlikely that these cells could be used for large-scale clinical myocardial replacement therapy.
A second potential source of cardiac cells is the resident pool of cardiac stem cells recently described in the rat heart11. But here again, the issues to be addressed are many, including the confirmation that cardiac stem cells exist in the human heart, the means of localizing and harvesting them noninvasively, their in vitro scalability or, alternatively, the identification of the signals required to recruit them endogenously in a regulatable fashion.
The limitations of both fetal and progenitor cardiac cells highlight the potential of a third source of cardiac cells: cells derived from pluripotent ES cells. Kehat et al. assessed the capacity of cardiomyocytes derived from human ES cells to function as pacemaker cells in vivo and thus to reestablish an electrical network in a heart deprived of its normal conduction system. Initial coculture experiments demonstrated a tight electromechanical coupling between the human cardiomyocytes and rat neonatal ventricular myocytes. This coupling seemed to involve gap junctions, as suggested by the immunostaining of connexin-43 at the interface between the two cell types and abolishment of conduction by the gap-junction uncoupler heptanol.
Importantly, the coculture data were confirmed by in vivo experiments in which ES cell−derived embryoid bodies transplanted into the posterolateral wall of a swine heart were shown to elicit, in half of the animals, a regular rhythm compatible with maintained hemodynamics following interruption of the normal atrioventricular pathway. Extensive ventricular mapping could localize the site of earliest activation in the transplanted area, with an excellent correlation between electrophysiological and pathological findings.
As acknowledged by the authors, a limitation of the study intrinsic to its design is that whole-organ mapping precluded the demonstration that electrical activity originated directly from the grafted cells. Likewise, an assessment of the cells' effects on left-ventricular function was not relevant to this noninfarction model. An additional limitation of the study is that characterization of the cardiac phenotype of the differentiated ES cells was based solely on immunostaining, which, in the absence of genetic markers, did not allow the question of fusion to be addressed. Finally, the cells were implanted in a noninfarcted tissue, which should optimize graft vascularization. Unfortunately, markers for cell proliferation, like Ki 67 labeling of the transplanted cells, were not assessed.
In spite of these limitations, the data presented in this study, taken collectively, provide fairly convincing evidence that human ES cells can electrically integrate in the recipient myocardium, suggesting that they could synchronously contribute to the augmentation of pump function following an ischemic injury. Importantly, the expected benefits of the electrical integration of the ES cell−derived graft were not offset by the development of teratomas or rejection (with the caveat that the animals were immunosuppressed).
Indeed, earlier studies conducted in rat models of myocardial infarction convincingly showed that, following appropriate in vitro precommitment towards a cardiomyogenic pathway, mouse ES cell transplanted into postinfarction myocardial scars could proliferate, differentiate into cardiomyocytes, survive for extended periods of time and improve left-ventricular function12. Kehat et al. extend these observations in a large animal model of xenotransplantation by documenting an efficacious pacemaker function of the grafted human ES cells.
The observation that the resulting idioventricular rhythm was responsive to adrenergic stimulation leads the authors to speculate about the cells' potential to act as a rate-responsive biological pacemaker. However, the remarkable safety and efficacy records achieved by current electronic pacemakers make it unlikely that we would prefer to rely on a cluster of grafted cells to take over a failing cardiac rhythm, and such an approach would probably not be compatible with the ever-growing safety considerations associated with any new treatment.
Furthermore, from the perspective of the treatment of arrythmias, it is likely that transplantation of ES cells would be targeted primarily at repopulating the sino-nodal area for treatment of atrial conduction disturbances, rather than the His bundle (the structure that conducts the electrical impulse between the atria and the ventricles), whose degeneration represents the primary indication for pacemaker implantation. For these reasons, the data presented in this paper should rather be viewed as an additional piece of evidence that human ES cells might be able to relieve heart failure by electrically coupling with the residual host cardiomyocytes and establishing a functional syncytium with them.
Notwithstanding the political and religious issues that unfortunately tend to cloud the debate, many technical obstacles remain to be surmounted before ES cells can be considered for safe clinical use, particularly the optimization of the cardiac precommitment and cell propagation procedures, the selection of those cells already engaged in the cardiomyogenic pathway to avoid in vivo uncontrolled proliferation and subsequent teratoma, the means of driving the transplanted ES cells towards complete maturation (instead of the embryonic-like phenotype used by Kehat et al.) to take full advantage of their contractile properties, the control of possible immune reactions and the long-term fate of ES cells engrafted in an infarcted tissue. Nevertheless, the apparent failure of adult stem cells to convert into cardiomyocytes, at least to a functionally relevant extent, strongly suggests that the ambitious goal of cardiac regeneration may ultimately be achieved only with ES cells. In this regard, the paper by Kehat et al. should be viewed as an important contribution.
A Persian proverb states: "Where there is a will, there is a path". We have the will. Let's find the path.
